[
https://issues.apache.org/jira/browse/MATH-1551?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17167935#comment-17167935
]
Gilles Sadowski commented on MATH-1551:
---------------------------------------
bq. push it yesterday.
I didn't see that.
I've pulled PR#157 right now, and it is not up-to-date with "master". You
should perform a
{noformat}
$ git rebase master
{noformat}
bq. there is a small deviation between the result from non-weighted percentile
and weighted percentile with all weights equal to 1. [...] in MATH-1491 [it
was] suggested adding tolerance to assertEquals when testing.
There, the error was at, or very close to, floating-point accuracy. While
here, one can suspect that the computation accumulates floating-point errors.
[You could confirm that by testing longer arrays.]
If that's the case, I suggest that you post to the "dev" ML, where you might
get suggestions on how this particular situation should be handled (there are
tools in "Commons Numbers" that could help).
> Compute Percentile with Weighted Samples
> ----------------------------------------
>
> Key: MATH-1551
> URL: https://issues.apache.org/jira/browse/MATH-1551
> Project: Commons Math
> Issue Type: New Feature
> Affects Versions: 4.0
> Reporter: institute for information industry
> Priority: Major
> Labels: features, newbie
> Fix For: 4.0
>
> Attachments: Percentile.java
>
> Time Spent: 40m
> Remaining Estimate: 0h
>
> The class Percentile only support estimation on non-weighted samples.
> I've implement some function to estimate percentiles of weighted samples.
> here is the reference:
> [https://stats.stackexchange.com/questions/13169/defining-quantiles-over-a-weighted-sample]!https://mail.google.com/mail/u/0?ui=2&ik=6059a2f2a6&attid=0.1&permmsgid=msg-a:r-8896240918589631988&th=1737bdaafc3e228f&view=fimg&sz=s0-l75-ft&attbid=ANGjdJ_GwCQ-hqi0o7ZJTEqcl6JLYbFiR2Y1sqxPL8jDNzheraAkX0beEUtAM4BYI2v_5XGvzCPj2gTBQutHT9bF8hYh7MgdiGYZhn0HtDuVswuCSuwy8aJVsvy3EqI&disp=emb&realattid=ii_kcyrjgea0!
>
> When all weights are equal to each other, it works like estimation on
> non-weighted samples under R-7.
> I can't find formulas for other rules but at least, now it can evaluate
> percentile for weighted samples.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
